Re: [Condor-users] hawkeye on dual-processor nodes

[Junjun Mao <jmao@xxxxxxxxxxxxxxxxx>]

Hi all,

A serious problem just happened to my cluster, causing entire shutdown 
of condor. The ownership of schedd was was changed to a regular user!!! 
How could this happen?

Here is condor related jobs left on the master node which is the submit 
machine.
[root@master1 y-61.1]# ps -ef | grep condor
pwang    26763     1  0 Nov18 ?        00:00:00 condor_shadow -f 886.0 
<10.10.20.1:34661> -
pwang    26766     1  0 Nov18 ?        00:00:00 condor_shadow -f 886.2 
<10.10.20.1:34661> -
pwang    26772     1  0 Nov18 ?        00:00:00 condor_shadow -f 886.1 
<10.10.20.1:34661> -
pwang    29394     1  0 Nov18 ?        00:00:00 condor_shadow -f 886.4 
<10.10.20.1:34661> -
condor   19319     1  0 Nov21 ?        
00:34:54 /home2/condor/sbin/condor_master
condor   19320 19319  0 Nov21 ?        01:43:02 condor_collector -f
pwang    19393 19319  0 Dec09 ?        00:00:06 condor_schedd -f
condor   19401 19319  0 Dec09 ?        00:02:31 condor_negotiator -f


Restarting condor daemons still yealds wrong owner of schedd. I have to 
move job_queue.log to another location to start condor correctly.

Can someone tell me where to look for the cause of the problem?

Junjun